Despite extensive research and intervention efforts the percentage of overweight Americans has continued to climb, putting a large segment of the population at risk for chronic disease. One reason for this failure may be that traditional research is not well suited for tackling the dynamic complexity of factors that result in a person becoming overweight. The present study takes up this challenge and adopts a dynamic systems approach to test a model that predicts the combined effects of relationship dynamics, emotions, and autonomic physiology on eating, activity, and short term weight gain in newly formed families. During the early years of marriage couples establish shared eating and activity habits that will subsequently be passed on to their children. Interrupting the obesity epidemic requires research focused on how couples develop shared obesity relevant behaviors. These mechanisms are amenable to behavioral intervention, but a key challenge for developing such interventions is our lack of understanding about how these factors interact. The basic premise of our model is that couple-level behavioral and emotional patterns focused around eating and activity provides proximal forces perpetuating eating and activity habits, with accompanying effects on body weight. We focus on two patterns that research on smoking and alcohol addiction suggests are important for maintaining unhealthy behaviors. The first, System-Symptom Fit (SSF), refers to an unhealthy behavior, such as excessive eating, helping to preserve relationship well-being by increasing positive emotion or couple closeness (Rohrbaugh, Shoham, Butler, Hasler, & Berman, 2009; Shoham, Butler, Rohrbaugh, & Trost, 2007). The second pattern, Demand-Withdraw (DW), occurs when one partner demands change in the other partner's health behaviors, which causes the nagged partner to withdraw from the interaction and to resist change (Eldridge, Sevier, Jones, Atkins, & Christensen, 2007). We propose that the presence of SSF or DW in newly formed couples will predict increases in weight and deteriorating health habits 6 months later. To test our hypotheses we will recruit 100 heterosexual couples who are within the first year of cohabiting. All participants will be in the upper-end of the healthy weight range at the beginning of the study. As such, they will be at risk of becoming overweight. Multi-method assessments of SSF and DW at baseline will be used to predict changes in eating/activity habits and weight 6 months later. This project has the potential for broad impact due to integrating across traditionally separate scientific and clinical communities, and by articulating a dynamic systems approach to the problem of obesity. As such, it could provide a theoretical basis for researchers and clinicians from diverse backgrounds. In addition, results from the study would inform research and practice regarding optimal, targeted ways to integrate family systems, emotion regulation, and biofeedback approaches to understanding and treating obesity.